load > pv during day

mevenable

I have a system with fifteen 270 W panels (4 kW), an Outback FM80 charge controller, GS8048 Inverter, a Mate3S control panel, and two strings of four NC200 batteries.  The system works great for my household needs, but I want to pump water from my well for the household and light irrigation, which would take about an hour or two in the middle of the day.  The pump is about 16-17 amps and can run at 14 amps with less flow, but with inefficiency and so forth about 0.5 to 1 kWH over the course of the pumping will have to come from the batteries.  The batteries could easily recharge on Flotation before the end of the day.  I am just not sure if the batteries are designed to handle that, or whether it will weaken them or shorten their life.  I am trying to get a third string of batteries, would that make my system stronger for pumping at a deficit in the middle of the day?  Or is it necessary to buy more panels (and another charge controller, since the FM80 is maxed out with 15 panels) in order to pump without messing up my batteries?

mcgivor

Welcome to the forum.

As long as the PV is able to support the majority of the pumps demand, which it appears to be, using 15A at 120Vac and 240Vac. Even at the higher voltage the pump would draw ~ 3.6Kw, a 4kw array would likely output ~3Kw so the net amount drawn from the batteries would be 600w, there is also power factor considerations which are unknown. If after pumping the battery voltage recovers rapidly to what it was prior to pumping there should be little cause for concern, taking a DC current measurement of what the inverter is using and compare this to what the array is producing, would be a more accurate means of gauging which is higher because it takes everything into account without having to calculate power factor, inverter losses, wiring losses and so forth. Can you fill the missing information battery voltage, pump voltage this would help in getting a clearer picture of exactly what it is you have, which would allow others to comment based on information rather than speculation.

Photowhit

Lots of things to discuss here. Your 2 strings of NC200 batteries would represent a 48 volt system. On a 48 volt system a single FM80 should be able to handle the charging from 4050 watt array by it's self.

Typically a 4050 watt array will produce 3010 watts in direct sunlight. 

So your pump uses 17 amps, I'll assume that's at 240 volts? for 4000 watts, So you will be drawing about 1000 + 300 from inverter inefficiency for 1300 watts from the battery bank assuming no additional loads. That shouldn't be a problem. It would be like running a toaster or microwave at night when you have no additional charging.

You are already slightly over paneled for a 400 amp hour battery bank being able to supply about 15% to a weak battery and 13% for a 50-25% discharged battery. I don't like more than 2 strings of batteries. I think you have a fine balanced system.

If you do add batteries, I would also add to your array unless you are in a particularly sunny area of the country (Southwest)

Editted because I read you already had 2 FM80's, Dang I'm getting old.

mike95490

With a 3Kw array, I used a ordinary garden timer to control my pump contactor, so it's only enabled 11am - 3:30pm

mevenable

Hi guys - thanks for the feedback so far.  The pump is 240 V and there are 2 strings of 4 batteries, 12 v each battery, total AH capacity of 400.  The solar array gives me 3.4-3.55 KW between 11 and 1 pm.  The pump appears to be needing 3.4 kW (14 amp use) and 3.8 kW (16.5 A use).  However, it almost certainly takes more than that, there are some other factors I am not sure of, but I have noticed that the higher the load, the more "excess" needed to handle it from the panels. Anyhow, the battery voltage goes down to 50.4 V when pumping, had been floating at 52.8-53.2 (set at 54.4 but maybe the lower voltage is due to temperature comp?  batteries are at about 31C/88F).  Anyhow, they eventually go down to 50 V then sometimes even 49.6 V or more if pumping at full capacity.  However, at the end I removed all load and charging, and the batteries settled in at about 50.8.  Then when charging, they float at 52.8 with a kW or so of input from solar almost immediately, and then it quickly goes down, and after an hour or two (depending on the deficit that I racked up) they are floating normally with 50 watts or so of input from the panels.  The reason I worry, is that I have been experimenting for a week or two, and the batteries started getting really weak at night, even if they seemed fully recovered hours before the sun went down.  The same accumulated load of about 2 KW hrs that used to bring them down to about 48.4 V by morning is now dumping them to the grid at 46V by 1 or 2 a.m.  They are less than half the strength they were, in spite of the fact I stopped pumping days ago and have avoided even using any larger loads at all.  They seem to be charging perfectly, very low amps by end of Absorption, low amps in flotation, all that.

mevenable

btw, I have one FM80, and from what I understand I can't add any more panels without another charge controller.  We pump into large tanks, which we then empty to drip irrigation system at night.  So, we are at present experimental stage very involved, although we could use timers eventually.  We only pump when the sun is strong, it is bright sunny all day since this is the dry season here (Central America) and we don't need to irrigate in the rainy season.  The Absorption generally ends by 10:30 or 11, takes about 2 hours, and then a float for about a half hour then pump

mevenable

oops - my bad, the batteries are 106NC, I have two strings and total of 200 AH -

mcgivor

mevenable said:

Hi guys - thanks for the feedback so far.  The pump is 240 V and there are 2 strings of 4 batteries, 12 v each battery, total AH capacity of 400.  The solar array gives me 3.4-3.55 KW between 11 and 1 pm.  The pump appears to be needing 3.4 kW (14 amp use) and 3.8 kW (16.5 A use).  However, it almost certainly takes more than that, there are some other factors I am not sure of, but I have noticed that the higher the load, the more "excess" needed to handle it from the panels. Anyhow, the battery voltage goes down to 50.4 V when pumping, had been floating at 52.8-53.2 (set at 54.4 but maybe the lower voltage is due to temperature comp?  batteries are at about 31C/88F).  Anyhow, they eventually go down to 50 V then sometimes even 49.6 V or more if pumping at full capacity.  However, at the end I removed all load and charging, and the batteries settled in at about 50.8.  Then when charging, they float at 52.8 with a kW or so of input from solar almost immediately, and then it quickly goes down, and after an hour or two (depending on the deficit that I racked up) they are floating normally with 50 watts or so of input from the panels.  The reason I worry, is that I have been experimenting for a week or two, and the batteries started getting really weak at night, even if they seemed fully recovered hours before the sun went down.  The same accumulated load of about 2 KW hrs that used to bring them down to about 48.4 V by morning is now dumping them to the grid at 46V by 1 or 2 a.m.  They are less than half the strength they were, in spite of the fact I stopped pumping days ago and have avoided even using any larger loads at all.  They seem to be charging perfectly, very low amps by end of Absorption, low amps in flotation, all that.

You have grid power? There is no sense in attempting to compete with the grid using battery storage, It will in most cases result in a net loss. The grid is, for no better term, raping your storage, if it is discharging the batteries overnight. Can you elaborate on what the complete system is, including what it is you are hoping to achieve? 

Estragon

If I'm understanding correctly, the system looks something like:
- 4kw panels, likely producing ~3kw when warm in full noon sun, assuming clean, properly tilted, and NO shade (not even nearby wires, pipes, etc).
- 3kw / 54v charging = ~56a
- 17a load at 240v ÷ .85 typical peak inverter efficiency = 4.8kw ÷ 50v = 96adc load.
- battery bank is 212ah@48v nominal. Note that 212ah is at 100hr rate.

In an off-grid system, the pv should provide 56a of the 96a load, with 40a coming from bank, so more like a C/4 rate (dropping effective capacity).

A couple of questions/suggestions occur. First, as the inverter is a "grid interactive" is it connected to grid? Is it properly configured? Others familiar with the Radian series may be able to help on this. My Outbacks just pass through grid/generator power, but Radians are different. Second, what size and lengths are the wires, particularly battery interconnects and controller and inverter wiring carrying high current lower voltage dc. How are the panels wired (5 strings of 3 in series?), how far away and what size wire? Are all the batteries a consistent temp (ideally measured with an ir thermometer)? Are any connections warm?

You mentioned the batteries "getting weak" overnight. What does "dumping them to the grid at 46v by 1am" mean? Is the inverter selling battery power overnight?

mevenable

Thanks again guys, much appreciated.  Actually, most of the time I am able to work at the 14 A setting, which APPEARS to need 3.4 kW.  The sun IS giving us about 3.45-3.55 kW, it is very sunny, the panels are optimally tilted, it is March, and we are a lot closer to the Equator than Arizona - 12 N.  I only need to use 3.8 kW when filling a certain tank that is a bit higher, and that takes about 15 minutes.  Some days I just used the grid for that one to avoid so much stress

There is a grid, but it goes down constantly, and no power means no water.  We first want to power the house consistently, and also 'go green" to some extent, so we set it up to mostly just cover our needs completely, which it was doing throughout February. We put the house on a large storage tank, small 1/2 HP pump easily handled by solar and battery, and pressure tank. However, for a relatively modest investment we could ramp up to the point that it would allow us to pump water from the well if the grid was down for days.  Having done that we were experimenting to see if we could save some money (grid power is very expensive here) by pumping water at midday to our large storage tanks for drip irrigation - basically become grid independent, although it would be there for backup.  Our Radian has a feature that when the batteries get low it will go to grid.  Supposedly you should not let them get much below 50% which I figure at 45-46 volts, so I set the thing to switch to grid when the batteries go down to 46.  That way, also, if there is a grid failure after that, before the sun comes up, they have a bit of juice left and wont get totally trashed.  So, this was never a problem - all through February the consumption at night barely got the batteries down to about 48.4 V, +/- 0.4.  The past week or so, the same consumption or less has taken the batteries down to 46 V and switched to grid.  This is in spite of the fact that it is charging fully (so far as can be seen) and we have not tried pumping for days.  I am currently running a full test (using for 10 hours with no pv, then cutting load and letting recover for 12 hours to check voltage.  These batteries are kind of weird btw - When the load goes off, they rebound a little, like from 46 to 46.8 V, in the first 20 minutes, then stay put for 3-8 hours, and then suddenly shoot up 2 V to 48.8V. - I did the test once before, and sometimes they do it in the early morning after having gone to grid

I think things are properly configured, I worked with the local solar company and with outback, and read every one of the manuals several times. I am not very sure of all the wire questions, it was installed by professionals and actually inspected by a trainer from outback who was here in January.  The panels are on the second story roof, and the wires go straight down from there to the rest of it, and the batteries are all together on racks with the most minimal cable connecting them, and everything else is within a meter or two.

The panels are 5 strings of 3

the connections and batteries do not seem warm, all charge equally, but one battery is producing some blue-green corrosion on the terminal from time to time.  I checked all the terminals with an ohm meter, and even the one with corrosion did not show resistance when the corrosion was cleaned off with baking soda

I really am thinking of ditching the NC batteries and just getting some golf cart batteries, there are some 12 V 150 A batteries available locally, but it is a shame that these NC106s have only been in use for 3.5 months.  I had issues in January, but the extra panels solved that, the batteries were charging fully even when cloudy after that

mevenable

ps, my peak inverter efficiency is supposedly 93%

Estragon

I don't know enough about the radian inverters to be much help, but from what you've said I'd lean towards a configuration issue.

Fully charged, the bank should have a resting voltage of ~51v. A bank that sits at 46v for hours with no load is substantially discharged. Shooting up to 48.x volts after sitting for 8 hours implies either a charging source or a bad connection getting good.

With the caveat that I don't have first-hand experience with radians, I would set it to switch to grid at ~51v unless grid power costs well over $1/kwh.

I'm also still not sure that the inverter isn't selling battery power into the grid at night. Before replacing the bank, I'd suggest a close look at inverter configuration. FWIW.

Raj174

@mevenable
Is this inverter a Radian GS8048A? If so, the Radian inverter could be set to enter Grid Zero mode before the pump starts. This will allow you to set the amount of power you want the batteries to supply and the grid will supply any amount over that. This will take the surge and highest part of the load off the battery bank, allowing a more reasonable amperage draw on the bank. Good and sometimes necessary for smaller battery banks.

This is done by setting the DOD amps setting in grid zero. This setting is in amps AC at 240 volts. For example, if you set "Grid Zero DOD Amps" to 8, this will supply about 1920 watts to the load and about 38 amps will be drawn from the battery bank and the rest will be supplied by the grid. The load on the battery bank can be set to what ever you feel comfortable with by adjusting the DOD amps setting. The switch from Mini Grid to Grid Zero mode and back, can be scheduled to occur at the same time daily, for any duration needed. Let me know if you need assistance with scheduling.

Grid Zero DOD Amps can be set with the mate 3, but is easier in OpticsRe. Open OpticsRe, select Device map, select the inverter, then select the battery charging tab. At the bottom of the page you will find the Grid Zero settings. Set DOD amps. Make sure the DOD volts setting is set low enough to supply the amps needed, I would guess 46 to 48 volts. If the DOD volts setting is too high, the inverter will not allow the voltage to drop below this setting and more of the power will be supplied from the grid. 

Next select the AC Input tab at the top of the page, scroll down the page util you see the word "Grid" on the left. Set "Grid Input Mode" to Grid Zero, then select the apply button at the bottom of the page. 
If the grid fails while the inverter is in the mode, the inverter will default to backup mode, and carry the load until the grid returns, then it should reenter grid zero mode. 

Rick
 

Photowhit

mevenable said:
.  Supposedly you should not let them get much below 50% which I figure at 45-46 volts, so I set the thing to switch to grid when the batteries go down to 46. 

They are AGM batteries, so hard to tell from voltage, but voltage at rest They will be around 50% capacity at 48 volts. At 45-6 volts you may well be damaging the batteries, if this is a resting voltage.

I mentioned  balance system earlier, if you have 2 strings at 100 amps or a 200ah 48 volt battery and a 4000 watt array which can deliver 3000 watts an hour or roughly 60 amps of 30% of the battery capacity. You have a somewhat unbalanced system. These AGM batteries might be able to handle it, but check the charge rate that is possible. This too, might be damaging your battery. 

What type of loads do you have over night?

noradawn456

Solar PV relies on light not heat. Many people think that a hot sunny day will produce more electricity than a colder sunny day. This just isn't the case. Solar PV relies on the light of the sun, but higher temperatures actually reduce the effectiveness of the panels. 

mevenable

45-46 V is with load of 1-2 amps, not at rest, after it switches to grid they rebound to about 47.2 V.  Our nightly load is commonly 2-3 kwh, and when the batteries are behaving well, that only takes them down to about 48 V at most with load. The charge rate is a good thing to consider, I could throttle that back so it does not overdo on a sunny day (where do I find that?), however, really the batteries go into absorption by 8:30 or so if sunny, so the maximum kW from the panels is about 2.2 at that hour, and then the charge controller takes over during absorption and flotation, it goes down pretty quickly once it enters absorption.  It would not damage the batteries if a very large amount of input from the panels was going to a load would it?  The grid zero thing sounds really interesting, I could never really figure out that setting but you have explained it well.  So I would be getting for example half of the load from the grid and half from the batteries?  at the same time?  very cool!  I am also thinking that the money I was going to spend on another string of batteries could better be put into another 6 panels and charge controller.  I would not have to have these extra panels even switched on except when I am pumping, or have a heavily clouded day.  Thanks everyone!  Btw I ran another one of those tests for outback, and so far the batteries ran for 6.5 hours then switched to grid due to low voltage, after just 1.9 kwh, and the battery has been resting at 46.8 V for the past 9 hours.

Estragon

With resting voltage at 46.8, the bank is likely around 30% SOC. At the relatively high rate of discharge, apparent capacity might be reduced to 170ah or so. 170ah x 48v nominal = 8.2kwh estimated capacity.

1.9kwh ÷ 70% DOD = 2.7kwh capacity as tested. Assuming the load was based on inverter output, the DC usage would be a bit more. Still, based on this test, the bank looks like it's lost a fair bit of capacity. This also fits with hitting absorb voltage at 8:30 with max pv rate of 2.2kw.

It may be worth checking individual battery voltages while resting. There's a chance damage might be limited to a single battery (eg one may have heated and vented gas), or one string (different resistance in wiring and/or batteries). Unfortunately, IMHO the most likely diagnosis based on information given is sulfation. This may be somewhat reversible with extended charging of individual batteries, but can be tricky with sealed batteries. Outback may be able to advise further.

mevenable

yow - weird!!!  the two strings are identical, with 46.8.  However, all the individual batteries on the lower string are reading 12.3 more or less, while the batteries on the upper string are measuring about 11.7 as might be expected.  How does the lower string DO that?  btw, the log files showed an average inverter current of 1 amp over the 6.5 hours the batteries held up during the test - the optics site shows 0.1 to 0.4 kW per hour.  Not high discharge rate, really

mevenable

oh shoot - that would be 2 amp average inverter current - every other space in the log is blank. still not so high, is it?

mevenable

now is charging and situation is the reverse - the top rack is still fine, the bottom rack has the same voltage across the string as the top, but the individual batteries all have voltages 1.7 V below what they should be

Photowhit

mevenable said:

yow - weird!!!  the two strings are identical, with 46.8.  However, all the individual batteries on the lower string are reading 12.3 more or less, while the batteries on the upper string are measuring about 11.7 as might be expected.  How does the lower string DO that?  

It's very difficult to maintain equal resistance over multiple strings. Setting them up with equal resistance, length and thickness of wire helps, but each connection is a possible point where a loose bolt might create problems. As I suggested earlier you may have damaged a battery/cell.

You might now did connect the weak/low voltage string and let it sit for a while. The other string is likely being brought down as it tried to send current to the weak string to reach the same voltage. Once disconnected, you are likely to find some cells/batteries increase in voltage and 1 decreases...

Best of luck! Sounds like a new install, maybe you can get it replaced under warranty.

BB.

All of the voltages have to add up to the same bus voltage.

If it is not a low cell\battery voltage, then it is a voltage drop at one or more connections. Note that to understand if you have good or bad connections (and even batteries\cells), you want a substantial current flow. If you do not have current flow, you cannot find bad connections (high resistance... V=I*R no current, no voltage drop).

Bill

jonr

>  46.8.  However, all the individual batteries on the lower string are reading 12.3

Impossible.  The sum of the voltages must equal the total.   But check each battery voltage for imbalance and also get a clamp on ammeter and check for balance from string to string.

BB.

Also check\change your DMM battery. I have seen near dead batteries give sort of repeatable readings that are wrong.

Bill

Estragon

> @mevenable said:
> oh shoot - that would be 2 amp average inverter current - every other space in the log is blank. still not so high, is it?

Not really, although not adjusting for discharge rate increases estimated capacity, so the loss of capacity vs tested just looks worse. I'd agree with Photowhit - split the two strings and charge individually, starting with the weaker one, and get both to the same (full) SOC.

Sealed batteries can't really be equalized, but it may be worth asking Outback if they can recommend a suitable sorta-EQ.

mevenable

Impossible but true - the lower string constantly gave the same overall reading as the upper, but the individual voltages were too high at rest, then too low when charging - i.e. they never moved, stayed at 12.3 V each more or less, when the string voltage indicated they should have been at 11.7 at rest and 13.3 while charging, then 13.9 while charging.  this was done in multiple iterations at widely spaced times with two different volt meters and two different people.  But then, as the batteries were absorbing with about 0.5 kW from the panels, suddenly it shot up to 2.2 kW from the panels, and . . . sure enough. . . the lower string finally started charging.  I would not have believed it if I had not seen it, but it also explains the day that the batteries were resting and the voltage suddenly shot up from 47.2 to 48.8 - the bottom string was not contributing all night, then suddenly decided to chip in.  Next thing is to figure out why, and what to do, but I have had the batteries for just 3.5 months, so I would say it could be a warranty issue.  I will let you know how Outback and the local company that deals their stuff respond, but they have been really good so far.  Except that they kept trying to tell me that having my voltages go up and down like a yo-yo was all normal.

BB.

Get a DC current clamp DMM. You need to measure current flow through each battery string.

It sounds like you may have an intermittent connection to the lower battery string. Open connection, no current flow, batteries don't change voltage, battery voltages do not add up to battery bus voltage, and your bank has only one half the storage capacity.

Bill

Photowhit

I made the poor assumption that the batteries had been disconnected for the measurement. 

If you have a fully charged string and an under charged string and you connect them + to + and - to -. You should have a 'system' and there should be flow from the fully charged string to the weak string.

I made the assumption that they had been disconnected but not at rest. I would like to see voltages with the batteries at rest.  As Bill said they may not drop in voltage, I suspect they will lose a surface charge. If yoou want to check apply a 12 volt load to each battery and watch voltage.

Are the strings connected correctly so that if you follow the flow of current that they pass through the same length and gauge of wire through each string?

Photowhit

Photowhit said:

If you have a fully charged string and an under charged string and you connect them + to + and - to -. You should have a 'system' and there should be flow from the fully charged string to the weak string.

In essence the weak string should be a constant load to the stronger string until they are at the same voltage. The reason you are seeing the weak morning battery may be the large difference. I still suggest as above, pull them apart and check them.

mcgivor

The 2 strings are part of the battery as a whole if, one string is of lower capacity, or suspected lower capacity, I would agree with removing it and testing each battery by fully charging letting it rest then read the voltage. Then apply a load to remove surface charge whilst monitoring voltage, if one or more cascade then those will be the ones in trouble. Another indication of lost capacity will be a rapid rise in voltage in an otherwise discharged battery once a charge is applied. Take notice of the battery temperature, reducing the charge voltage accordingly by subtracting 0.15V for every 5°C above  25°C of the manufacturers recommendations. If there is one bad cell with a low voltage  the remaining cells will rise in voltage to a point of overcharge, while the battery voltage as a whole appears normal, not a good thing with sealed batteries. With flooded it would be as simple as taking SG readings, unfortunately this is not an option, so voltage and load testing is about all one could do.
The one outstanding comment was the battery being in the 46V range, you really don't want to go much below 48V on a regular basis to avoid problems. My opinions and thoughts.

mevenable

ok, the batteries are connected, yes, and properly connected, by the company that installed them,  I would love to disconnect them and test them and do all that, but it is more than I am personally able to do, for a number of reasons,  I will see if the company can come by and do all that.  Two times I tested the strings and individual batteries at rest, twice while they were charging.  While at rest, the individual batteries on the lower string were higher than they should have been (compared to the reading for the string, which was identical to the upper string), and when they were charging, they were lower, and in fact, the individual lower batteries were always the same, they were not charging or discharging.  Then 45 minutes into absorption, the amount of current from the panels went way up, from 0.5 kW to 2.2 kW delivery, and I realized that the lower batteries were starting to charge.  I measured them and sure enough.  For a while, the lower batteries were a bit lower than the upper ones, and then caught up, but the range of voltages was fairly high, even on the upper batteries, a difference of almost 0.5 V from the lowest voltage to the highest voltage.  Then by the end of absorption, they were all within 0.2 V of each other, which is what I have found generally when I measure them.  Normally, when the batteries are all engaged and behaving themselves, my nightly load does not take them below 48 V.  Last night 1.6 kWh brought them down to 48.4 V with a 0.2 kW load applied.  That is a bit weak but at least closer to normal, for a long while 2 to 2.5 kW took them down to 48.4.  Thanks guys, at least your advice gives me something to ask for.